Dear All,
After switching from 0.8.14 to 0.9.0 I have this problem of getting
error: 400; check if you use aliases in SER
when I try to add a user with serctl,
eg.serctl add 01548900 999 01548900(a)redtone.com
I searched the web and knew someone else has posted this question before
and I apologize for asking again ( didn't manage to get the answer).
Thanks for any help available.
Regards,
TC Chan
I fixed the problem commenting the revert_uri(), so it call each time the
modified uri that i have after lookup("aliases").
very easy after all.
What I am not able to preview is the future implication of commenting
revert_uri() giving the fact that in all samples it is present.
Thanks
Rosario
----- Original Message -----
From: "Stefan Sayer" <sayer(a)fokus.fraunhofer.de>
To: "Rosario Pingaro" <rpingar(a)italycom.it>
Sent: Thursday, June 16, 2005 6:49 PM
Subject: Re: Fw: [Serusers] voice mail problem on alias
> Hello,
>
> that is interesting - if you find the answer can you mail me as well?
> (I'd like to add that to the documentation.)
>
> I think it depends on how you read the email adress from the db - probably
> you are loading it with something like
> avp_db_load( "$ruri", "$email/$email_scheme");
> and if it is not found you'd have to check if it is an alias and lookup
> the real name.
>
> Another option is to use user_prefix_separator, only a single ser
> instance, I think with this scheme there is not this problem with the
> alias (but i am not sure).
>
> Stefan
> Rosario Pingaro wrote:
>> Debugging sems seems that the probelm is related on how ser pass the user
>> to sems;
>> in fact when I call the alias and the use is ofline ser pass the real
>> user to sems.
>> when ser get a timeout pass to sems the alias, sems is not able to find
>> an email for the alias and so i get the error.
>>
>> How to workaround this?
>>
>> Thanks
>> Rosario
>>
>>
>> ----- Original Message -----
>> From: Rosario Pingaro To: serusers(a)lists.iptel.org Sent: Thursday, June 16,
>> 2005 5:20 PM
>> Subject: [Serusers] voice mail problem on alias
>>
>>
>> Hi I am implementing voice mail using ser and sems, ser 0.9.2 and latesr
>> sems from cvs.
>>
>> I have the failure_route[x] to activate the voicemail also when i get an
>> invite timeoute for my users.
>> failure_route[1] {
>> revert_uri();
>> rewritehostport("xxx.xxx.xxx.xxx:5090");
>> append_branch();
>> t_relay_to_udp("xxx.xxx.xxx.xxx", "5090");
>> }
>>
>> I don't have any problem when i try to leave a voice mail for an "user".
>> On the "alias" the voicemail works fine when the user is offline, but
>> when it is online and get a timeout sems respond with a 404 no email
>> address for user
>>
>> the fact that the voicemail works on alias when it is off line push me to
>> think the problem is into the failure_route. But I don't know what to add
>> to get it working.
>>
>> Any help is welcome.
>>
>> Thanks
>> Rosario
>>
>>
>>
>> --------------------------------------------------------------------------------
>>
>>
>> _______________________________________________
>> Serusers mailing list
>> serusers(a)lists.iptel.org
>> http://lists.iptel.org/mailman/listinfo/serusers
>>
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Serusers mailing list
>> serusers(a)lists.iptel.org
>> http://lists.iptel.org/mailman/listinfo/serusers
>
>
Hi,
This is a notification that the first release of free-TLS for SER has
been commited to the experimental tree in the CVS repository for SER.
For openSER it will be (i think) committed to the core CVS repository
right away.
User testing is required to fast-track it into the main core, so I
encourage you to give it a try. More will be added soon to ease
testing (for example, test certificates ready to use ... but you can
also create your own).
It contains a README file, which should be enough to get things
started. It will be improved with a fully working tls_only.ser.cfg
file in no time. Also more developer help will be added in the README,
for people who want to collaborate.
This is all for now.
Regards,
Cesc
yes i read it -:)
I would have spent less time if i had read it but it's
not wasted time .
I use SER I have to read OPENSER docs.
Something is wrong !
Thanks all for help.
Harry
--- Bogdan-Andrei Iancu <bogdan(a)voice-system.ro> a
écrit :
> Harry,
>
> you may find this useful
>
> http://www.openser.org/docs/modules/domain.html
> also you can find full up2date docs for all modules
> there
>
> bogdan
>
> harry gaillac wrote:
>
> >According to domain.c file "is_uri_host_local()"
> >function
> >is needed to check R-URI.
> >
> >Harry
> >
> >--- Juha Heinanen <jh(a)tutpro.com> a écrit :
> >
> >
> >
> >>put xlog debug into the script. print from header
> >>before the tests,
> >>etc. also, wheck with fifo command domain_dump
> (or
> >>something like that)
> >>that domain table is really loaded.
> >>
> >>-- juha
> >>
> >>
> >>
>
___________________________________________________________________________
Appel audio GRATUIT partout dans le monde avec le nouveau Yahoo! Messenger
Téléchargez cette version sur http://fr.messenger.yahoo.com
Hello! I'm playing around with my own module which is adding or
replacing a header in proxied INVITE requests.
I'm using the "parse_headers ... anchor_lump ... insert_new_lump_before"
method on the proxied request and the outgoing INVITE looks
great with the new header.
But the request is forwarded by using the tm module and when the
transaction receives a CANCEL the outgoing CANCEL gets the
header added to the initial INVITE. That would not be a problem
for a normal SIP UA, since unknown headers should be ignored,
but this UA from a big telecom vendor throws a 400 Bad request.
Why is the added headers added to the CANCEL? Shouldn't just
the required and transaction identifying headers be enough in a
CANCEL? Could I do anything different to prevent this from
happening?
/Henrik Gustafsson
Hello All,
Looking for some input if anyone else has run into this situation
and/or figured a way around it.
- Call comes in from PSTN gateway to a local user
- SER finds the user in the location table and t_relay pushes the
request off to them
- The user is behind an ACL and is not sending keep-alives (yes I know
the client is broken) so the INVITE never makes it to them and hence no
180/183 Ringing message comes back
- PSTN user gives up and hangs up the phone after 3 seconds of no ringing
- PSTN gateway forwards the CANCEL message in and SER cancels the branch
- However since t_relay is stateful and re-tries transmissions the
INVITE is sent another time or two
- Finally the connection times out and the failure_route set for the
t_relay kicks in and the INVITE gets pushed off to the voicemail server.
- Voicemail server OK's the INVITE and starts retransmitting the OK over
and over and over since the PSTN GW doesn't ACK the OK since it knows
that call-id as being cancelled.
So, should I handle cancel's a little more explicitly by doing
something similar to:
if ((method=="CANCEL") && (t_lookup_request())) {
t_release();
};
What side effects, if any, would this cause? Is there a better way to
handle it?
Thanks,
-Evan
Folks,
There were a lot of ideas, discussions, fears vehiculated in the last
days. I try to bring some response as quick and as short as possible
just to close the subject.
1) there is and it will be no war - I think the idea to associate war
with new thing is in human nature and we cannot fight it. Media Proxy
was also welcomed with war and it proved to be a good thing (a
complementary solution to nathelper); avpops got the same treatment and
now I believe to be a very used module. So let's try t get rid of the
war idea - OpenSER is just an alternative to SER and everybody is free
to choose what to use.
2) very common idea (no idea who came with it?!) was that OpenSER is
less stable and it will be mainly an place for new thing to get mature -
wrong ! OpenSER's stability will be one of its strong attribute - if you
look to the NEW FEATURE section we will be surprise how many bugs were
hunted and fixed. Shortly - we will try to keep OpenSER as stable as SER
and maybe more - of course, nothing is bug free :)
3) getting back into SER team - maybe the GCC story will happen, but for
this the SER must prove (with facts and not words) that it became a real
open project - politics and development are not dictated by any hidden
commercial reasons; SER becomes an independent and neutral project (not
faked "propriety" of anybody), anybody being free to contribute; the
maintainer dictatorship must end - a module maintainer must not have
"live or death" right on its modules, it must obey the management group
of the project!.
4) documentation management - that a hard problem which frustrated the
users a lot - OpenSER tries a different cycle of documentation which
will make it easier to use and to update for user: main storage point
will be sgml files on CVS -> HTML files on web for reading -> on web
wiki page for user to contrib/comment -> filtered time to time by
maintainers and moved into sgml; and here ends; easy and straight for
both users and maintainers; also Makefiles has built-in commands for
generating READMEs and HTMLS from sgml.
now, some short answers to some interesting ideas:
Greger
------
what was the attitude when "commercial components start popping up with
alternative contributions" ?
IGNORE!
"Do they(voice system) contribute their commercial code to OpenSER? Of
course not, they have exactly the same problems as iptelorg.com" Maybe
you should look twice to avpops, speeddial, alias_db, uac modules,
maintenance for xlog and pdt and all patches to usrloc, rr , register
and nathlper - many enterprise solution fixes are there (nat traversal
and replication)
Kristian Galway, reticent, Maxim, Klaus, Giudace
------------------------------------------------
there is no war and let's forget the word! it's simple matter of
choosing and liberty (even if history showed us that all this were gain
through war) Everybody is free to use whatever he like. If you don't
like OpenSER, fine by me, forget about it, but don't misplace words!
Dave, Ingo, Maxim, Atle
------------------------------------------------
by splitting the project nothing is lost, but better (efficient) used;
for developers - no more struggle to put your stuff on cvs; more time to
work on real things; anyhow you worked was blocked on SER :(. For users
- if they know what they want, they can have the liberty to choose based
on the preferences, which is a step forward. Let's not hide behind the
mentality - simple thing and less options make our life easier...
Harry, Samuel
-------------
interesting idea of being looser....wonder way?? If like SER you can
stick to it and you don't loose or win; If you find OpenSER more
suitable for you, you win.....so....where is the looser may I ask??
Cesc
-----
|And this process (n. devel) must be well-defined and NOT CONTROLLED exclusively by a company (be iptel or voice-systems, same-o, same-o). If the contribution is voted as desirable by a community of users, it is well tested and so on, it should be accepted."
you made a strong point here !!
with this email I want to close the subject. Please *do not reply* to
it, but only in privately or use openser mailing list if you have
related questions.
End of subject
Bogdam
Does any one have a sample cfg for a ser + asterisk setup they would like to
share?
We are planning to use asterisk to provide vm, conferencing and pstn
functionality.
So any input on this issue would be nice.
Hello,
I have some more information about this issue. There are actually two problems.
First - When the 1st attempt fails because connection was refused, connection timed out, or there was no route to host, an error message as logged and the failure_route block is never executed.
Second - When the 2nd or 3rd, etc attempt fails due to a connection problem, an error is logged and only 30 seconds later the failure_route block is executed.
I would really appreciate if some body could explain what is going on or point me in the right direction.
Thank You,
Dmitry
----- Original Message -----
From: Dmitry Isakbayev
To: serusers(a)lists.iptel.org
Sent: Wednesday, June 15, 2005 9:12 PM
Subject: How to execute failure_route block on tcp connect failure.
Hello,
I am having a problem executing a failure_route block when t_relay fails to establish a tcp connection. Everything works as expected when t_relay connects to a proxy and receives back an error like 501. Is there a way to configure t_relay to execute a failure_route block an a tcp connection error?
Thank You in advance.
Dmitry
Here are the logs from connection failure -
Jun 15 20:52:14 localhost ser[21870]: ERROR: tcp_blocking_connect:
SO_ERROR (111) Connection refused
Jun 15 20:52:14 localhost ser[21870]: ERROR: tcpconn_connect:
tcp_blocking_connect failed
Jun 15 20:52:14 localhost ser[21870]: ERROR: tcp_send: connect failed
Jun 15 20:52:14 localhost ser[21870]: msg_send: ERROR: tcp_send failed
Jun 15 20:52:14 localhost ser[21870]: ERROR: t_forward_nonack: sending
request failed
Sorry, but I didnt explained my self corectly,
what I want is to redirect the calls that aren't for my domain ( for
example, calls to sip:maria@fwd.pulver.com or calls to sip:manel@iptel.org).
What I whant to know is what to put after this:
if( ! uri=~"@fccn.pt" ) {
#redirect to fwd or iptel
Thanks
João
Klaus Darilion wrote:
> The suggest way is ENUM, but only a few of them creates ENUM entries
> for there users.
>
> regards,
> klaus
>
> Joao Pereira wrote:
>
>> Hello to all
>> I whant my SIP clients to be able to do calls to any other SIP
>> network, and to do It, I writed this line (for the IPTEL.ORG case):
>>
>> if(uri=~"@iptel.org"){ forward(iptel.org,5060);
>> break;
>> }
>>
>> But this way, I need to put theese 4 lines to all VoIP providers.
>> Is there any way of saying:
>> if( ! uri=~"@fccn.pt" ) { # if the invite isnt for my domain (fccn.pt)
>> send it to the correct domain
>>
>> Thanks
>>
>> João Pereira
>>
>>
>>
>> _______________________________________________
>> Serusers mailing list
>> serusers(a)lists.iptel.org
>> http://lists.iptel.org/mailman/listinfo/serusers
>>
>>
>
>